Image Captioning Project CSC413

Overview

This repository contains the training code for a couple image-captioning models.
The Flickr8k dataset was used to train the models.
Validation and Evaluation code using BLEU.
The training and validation code is all original while the encoder decoder models were modified for the project and are from other repositories (links below)

Name		Name	Last commit message	Last commit date
Latest commit History 58 Commits
flickr8k		flickr8k
models		models
models_without_attention		models_without_attention
AlexnetModel.ipynb		AlexnetModel.ipynb
README.md		README.md
data.ipynb		data.ipynb
data.py		data.py
encoder_decoder.py		encoder_decoder.py
main.ipynb		main.ipynb
splitdata.ipynb		splitdata.ipynb
train.py		train.py
transformer.py		transformer.py
utils.py		utils.py
validation.py		validation.py
validation_test.ipynb		validation_test.ipynb
vocab.pkl		vocab.pkl
vocab.py		vocab.py